http://bugzilla.opensuse.org/show_bug.cgi?id=1183839 http://bugzilla.opensuse.org/show_bug.cgi?id=1183839#c2 --- Comment #2 from Bruno Friedmann --- Created attachment 847491 --> http://bugzilla.opensuse.org/attachment.cgi?id=847491&action=edit RIP: 0010:kobject_put+0x19/0x1d0 kernel with 5.11.6 (without nvidia) This time I've tried to not taint the kernel by removing completely the nvidia drivers, but the crash still occurs. -- You are receiving this mail because: You are the assignee for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1183839 http://bugzilla.opensuse.org/show_bug.cgi?id=1183839#c3 --- Comment #3 from Bruno Friedmann --- Created attachment 847492 --> http://bugzilla.opensuse.org/attachment.cgi?id=847492&action=edit Dell Precison M7510 lshw Hardware description -- You are receiving this mail because: You are the assignee for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1183839 http://bugzilla.opensuse.org/show_bug.cgi?id=1183839#c5 --- Comment #5 from Bruno Friedmann --- (In reply to Takashi Iwai from comment #4)

As far as I see in the previous bug report, the first Oops is always same in this code path from various 5.11.x kernels, right?

And, you've tried to blacklist dell-wmi-sysman module but then you got a workqueue stall (in https://bugzilla.opensuse.org/show_bug.cgi?id=1182377#c29)?

The Oops itself looks like some memory corruption or such, which leads to the next crashes. But, even if the problem is triggered without dell-wmi-sysman, it means the corruption already occurred before that point.

Humm this sound perticulary bad. I believe I will have to open the lappy, and check all connections, and run the highly long full Dell hardware check (with 64GB it always takes time). I will report later the result. I've to move all the vm and container running on this datacenter :-) -- You are receiving this mail because: You are the assignee for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1183839 http://bugzilla.opensuse.org/show_bug.cgi?id=1183839#c7 --- Comment #7 from Takashi Iwai --- I meant a memory corruption by some bad kernel code, not the hardware defect. You can try to install kernel-debug and see whether this catches anything. Also, systemd.confirm_spwan=1 boot option will allow you step-by-step boot sequence. I guess it'll be the udev-trigger or such stage where everything happens, but it might narrow down a bit. Better to disable plymouth if you do that, though (via plymouth.enable=0 option). -- You are receiving this mail because: You are the assignee for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1183839 http://bugzilla.opensuse.org/show_bug.cgi?id=1183839#c9 --- Comment #9 from Bruno Friedmann --- Created attachment 847524 --> http://bugzilla.opensuse.org/attachment.cgi?id=847524&action=edit Picture of the Rip 2 If you have another option on boot that I can try (it seems I've not started the -debug kernel in my last attempt, so I will redo that shortly) options used nosplash silent plymouth.enabled=0 noresume crashkernel=256M-:128M rd.vconsole.font=ter-v32b.psfu rd.vconsole.keymap=ch-fr rd.locale.LANG=en_US.UTF-8 audit=0 apparmor=0 mitigations=auto nvme_core.default_ps_max_latency_us=5500 scsi_mod.use_blk_mq=1 usb_storage.quirks=0x0bc2:0xab38:u systemd.confirm_spwan=1 single -- You are receiving this mail because: You are the assignee for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1183839 http://bugzilla.opensuse.org/show_bug.cgi?id=1183839#c26 --- Comment #26 from Bruno Friedmann --- (In reply to Takashi Iwai from comment #25)

Do you still face the problem with latest upstream? i.e. the sound module is still included in initrd? If yes, we need to track down the cause.

Hi Takashi, I've just done a new test this morning with kernel 5.12.10-1-default and I still have some snd modules added to the initrd lsinitrd | grep snd -rw-r--r-- 1 root root 7908 Jun 15 18:32 usr/lib/modules/5.12.10-1-default/kernel/sound/core/snd-hwdep.ko.xz -rw-r--r-- 1 root root 66888 Jun 15 18:32 usr/lib/modules/5.12.10-1-default/kernel/sound/core/snd-pcm.ko.xz -rw-r--r-- 1 root root 18416 Jun 15 18:32 usr/lib/modules/5.12.10-1-default/kernel/sound/core/snd-rawmidi.ko.xz -rw-r--r-- 1 root root 5228 Jun 15 18:32 usr/lib/modules/5.12.10-1-default/kernel/sound/core/snd-seq-device.ko.xz -rw-r--r-- 1 root root 19364 Jun 15 18:32 usr/lib/modules/5.12.10-1-default/kernel/sound/core/snd-timer.ko.xz -rw-r--r-- 1 root root 42836 Jun 15 18:32 usr/lib/modules/5.12.10-1-default/kernel/sound/core/snd.ko.xz -rw-r--r-- 1 root root 42912 Jun 15 18:32 usr/lib/modules/5.12.10-1-default/kernel/sound/hda/snd-hda-core.ko.xz -rw-r--r-- 1 root root 5680 Jun 15 18:32 usr/lib/modules/5.12.10-1-default/kernel/sound/hda/snd-intel-dspcfg.ko.xz -rw-r--r-- 1 root root 3336 Jun 15 18:32 usr/lib/modules/5.12.10-1-default/kernel/sound/hda/snd-intel-sdw-acpi.ko.xz -rw-r--r-- 1 root root 72856 Jun 15 18:32 usr/lib/modules/5.12.10-1-default/kernel/sound/pci/hda/snd-hda-codec.ko.xz -rw-r--r-- 1 root root 27516 Jun 15 18:32 usr/lib/modules/5.12.10-1-default/kernel/sound/pci/hda/snd-hda-intel.ko.xz I'm normally not too much annoyed actually as I used to use a dracut configuration which force it to avoid adding them /etc/dracut.conf.d/80-no-sound-modules.conf omit_drivers+=" snd_hda_intel snd_seq_device snd_hwdep snd_hda_intel snd_usb_audio snd_usbmidi_lib snd_hda_codec snd_timer snd_compress snd_soc_core snd_pcm snd_rawmidi snd_intel_dspcfg soundwire_intel soundwire_intel snd_compress snd_soc_core snd_hda_core snd_pcm_dmaengine snd_soc_core snd_compress snd_seq_device snd_hwdep snd_timer snd_pcm snd_timer snd soundcore_bus soundcore soundwire_generic_allocation soundwire_bus soundwire_cadence " As extra modules I only have nvidia, and virtualbox (all coming from openSUSE official rpm). -- You are receiving this mail because: You are the assignee for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1183839 http://bugzilla.opensuse.org/show_bug.cgi?id=1183839#c28 --- Comment #28 from Bruno Friedmann --- (In reply to Takashi Iwai from comment #27)

OK, then could you try to uninstall Nvidia package once, rebuild initrd and check whether initrd still contains the sound module? If it's still there, uninstall vbox once and check again. If both drivers are uninstalled but still the sound module gets dragged, some other configuration must be the cause. Maybe better to ask dracut people for taking a look. A typical cause in that case is some leftover setup for the kernel.

Hello Takashi, sorry was off a few days. I've tried by uninstalling nvidia and virtualbox, and rebuild with my dracut exclusion configuration file. snd modules are present in the initrd. I was not yet able to disconnect everything (take the laptop out of the base without any additionnal peripheral like screens, webcam etc). Perhaps the snd come now from one of those, but the laptop serve here a central datacenter with a lot of vms so no easy to get out of order to make tests (sorry) I'm pretty sure it is not nvidia, nor virtualbox, and on another computer, with both installed snd doesn't appear on initrd. :-( Hopefully the trick of using an omit_driver with dracut works, and allow me to use again the computer. So in a sense it now works for me, we just don't know why it failed .... -- You are receiving this mail because: You are the assignee for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1183839 http://bugzilla.opensuse.org/show_bug.cgi?id=1183839#c30 Bruno Friedmann changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution|--- |FIXED --- Comment #30 from Bruno Friedmann --- (In reply to Takashi Iwai from comment #29)

I guess we have to check the dracut behavior step by step. My wild guess is some leftover configuration or setup file that matters; e.g. /usr/lib/modules-loaded.d/* or /etc/modules-loaded.d/*, also some /etc/sysconfig/*, whatever...

OMG I found a culprit in /etc/modules-load.d/ -rw-r--r-- 1 root root 14 Jan 24 2018 yast.conf cat /etc/modules-load.d/yast.conf snd-hda-intel Bingo! Removing this one just kill the problem, no more snd module in initrd You already know that, but you're an ace!!! I'm just closing now this report. Many����� Thanks -- You are receiving this mail because: You are the assignee for the bug.